ESPN raises statistical confusion to a new level

[Marcus92]

ESPN seems obsessed with: 1) creating new proprietary statistics; 2) making them impossible for anyone to understand; and 3) patting themselves on the back for how unbelievably awesome they are.

The latest is BPI (Basketball Power Index). It simplifies everything down to a single number, with nothing to explain their methodology or results, no supporting data, nothing to interpret. Instead, you get 1,000+ word articles like this:

http://www.espn.com/blog/statsinfo/post/_/id/131084/how-did-espns-upgraded-bpi-perform-this-college-basketball-season

Who's to say they didn't just pick a number out of thin air at the end of the season and say that BPI outperformed KenPom (supposedly in terms of predictability)? Statements such as "BPI preseason rankings are meant to predict what a team's BPI will be at the end of the season" say absolutely nothing.

What is ESPN really trying to accomplish here?

[GooooMarquette]

It's simple, really.

ESPN = mc²

[Dr. Blackheart]

ESPN no matta

[mu03eng]

ESPN seems obsessed with: 1) creating new proprietary statistics; 2) making them impossible for anyone to understand; and 3) patting themselves on the back for how unbelievably awesome they are.

The latest is BPI (Basketball Power Index). It simplifies everything down to a single number, with nothing to explain their methodology or results, no supporting data, nothing to interpret. Instead, you get 1,000+ word articles like this:

http://www.espn.com/blog/statsinfo/post/_/id/131084/how-did-espns-upgraded-bpi-perform-this-college-basketball-season

Who's to say they didn't just pick a number out of thin air at the end of the season and say that BPI outperformed KenPom (supposedly in terms of predictability)? Statements such as "BPI preseason rankings are meant to predict what a team's BPI will be at the end of the season" say absolutely nothing.

What is ESPN really trying to accomplish here?

Same thing they tried to do with QBR, trying to create an "universal" stat that all other competitors will have to reference to talk to customers which will in turn drive eyeballs to E$PN as the only true source of this wonderful "universal" stat

It's crap, KenPom rules foreva

[GGGG]

Same thing they tried to do with QBR, trying to create an "universal" stat that all other competitors will have to reference to talk to customers which will in turn drive eyeballs to E$PN as the only true source of this wonderful "universal" stat

It's crap, KenPom rules foreva

The one thing that QBR does is put everything on a 0-100 scale.  Versus 154.9 or wherever passer rating tops out.

[Jay Bee]

ESPN can't even calculate RPI. They probably screw up calcs of their own crap.

[avid1010]

Plenty of bias and curcular references in quantitative data...for the average fan...hard 2 know what is what.

[2012 Warrior]

The only worse 'statistical' ranking system is the Audi Player Index in the MLS... Just absolute crap

[Mr. Sand-Knit]

The only worse 'statistical' ranking system is the Audi Player Index in the MLS... Just absolute crap

Went to nap vs juve and later roma v lazio last week
Good time

[willie warrior]

It's simple, really.

ESPN = mc²

Even simpler. Espn=Disney PC BS.

[Newsdreams]

ELP>BPI

[Eldon]

Just to be sure, the BPI is at least 5 years old if not older, FWIW

[Marcus92]

Just to be sure, the BPI is at least 5 years old if not older, FWIW

Yeah, BPI isn't new — although they've apparently done something to upgrade/improve the system. I'm just continually amazed at how little worthwhile content ESPN offers on college basketball. Some sports leader.

[Marcus92]

ELP>BPI

El Paso Airport?

Extra Long Play?

Emerson, Lake & Palmer?

[#UnleashThePortal]

How does one patent math?

[Newsdreams]

El Paso Airport?

Extra Long Play?

Emerson, Lake & Palmer?

#3

[GooooMarquette]

#3

I agree with this analysis.

[TSmith34, Inc.]

Same thing they tried to do with QBR, trying to create an "universal" stat that all other competitors will have to reference to talk to customers which will in turn drive eyeballs to E$PN as the only true source of this wonderful "universal" stat

It's crap, KenPom rules foreva

It might very well be crap, but it actually slightly more logical than passer rating.  No one pays attention to it, and rightly so, but at least it tried to put the stats into perspective by weighting them to the situation, versus QB rating where garbage time weighs equally. 

I agree with *why* they came up with their own stat, but that doesn't mean it isn't actually better than the original passer rating.

[GB Warrior]

How does one patent math?

Isaac Newton and Leibniz want to know before ESPN can patent calculus

[muwarrior69]

The only stat that mattas is the W/L ratio. If it is grater than 1, your team is a winner.